ABSTRACT/SUMMARY ? Project 3: Integrative Computational Analysis of Tumor-Mediated Immunosuppression Lymph nodes are typically assessed in cancer patients to determine disease stage and treatment plan, yet they are poorly understood and largely understudied in the context of metastatic progression. Most public- domain high throughput cancer datasets, including The Cancer Genome Atlas (TCGA), profile the primary tumor and sometimes distant metastatic sites, but rarely positive lymph nodes. In this proposal, we turn our attention to the lymph nodes with a cancer systems biology approach that tests the hypothesis that lymph nodes metastasis establishes the gateway to distant metastases by enabling immune tolerance to tumor- associated antigens. Our Research Center will generate complex high dimensional datasets in both the primary tumor and lymph nodes of patients with melanoma and head and neck squamous cell carcinoma (HNSCC) to explore this hypothesis in both human and mouse studies. From each tissue specimen, in both the human and mouse, we will generate RNAseq data of sorted malignant and immune subpopulations (Projects 1, 3) and high dimensional in situ images at the level of single cells on the bulk tissue (Project 2). The goal of this Research Project is to develop and apply computational tools to integrate these complex datasets in order to identify candidate mediators of tumor-immune interactions that induce immunosuppression for functional validation. To enrich our ability for interpretation, we will explore signatures of the immune system in a pan-cancer analysis using the TCGA datasets annotated with time to distant metastasis, in the context of node-negative and node-positive patients. We hypothesize that pan-cancer genes whose expression is strongly associated with time to distant metastasis are more likely to be associated with tumor-intrinsic or microenvironmental processes driving metastasis progression; thus we will prioritize these genes in our integrative computational analysis of our melanoma and head and neck squamous cell carcinoma datasets. Using the RNAseq data generated by our study, we will develop and apply novel network-based computational methods for reconstructing the interactions between malignant and immune subpopulations. Moreover, we will develop and apply new approaches to integrate the spatial information from high dimensional single cell in situ images with the gene expression datasets to further refine our inferences of candidate mediators of immunosuppression. The datasets and computational resources developed by our Research Project, and Center at large, will not only enable use to deeply explore the role of lymph nodes in tumor-mediated immunosuppression, but will also provide the community with powerful resources for understanding systemic influences on the forces governing metastatic dissemination.